Repeat patient testing‐quality control with canine samples shows promise as an alternative to commercial quality control material for a network of four Sysmex XT‐2000iV hematology analyzers

Abstract

AbstractBackgroundRepeat patient testing‐quality control (RPT‐QC) uses retained patient samples as an alternative to commercial quality control material (QCM). We elected to calculate and validate RPT‐QC limits for red blood cell count (RBC), hemoglobin (HBG), hematocrit (HCT), and white blood cell count (WBC).Objectives(1) To validate RPT‐QC across a network of four harmonized Sysmex XT‐2000iV hematology analyzers and determine the total error that can be controlled with RPT‐QC. (2) To generate quality control (QC) limits using the standard deviation (SD) of the duplicate measurement differences and determine a suitable simple QC rule with a probability of error detection >0.85 and probability of false rejection <0.05. (3) Monitor RPT‐QC using sigma metrics as a performance indicator and (4) to challenge RPT‐QC to ensure acceptable sensitivity.MethodsFresh adult canine EDTA samples with results within reference intervals were selected and run again on days 2, 3, and 4. QC limits were generated from the SD of the duplicate measurement differences. The QC limits were challenged using interventions designed to promote unstable system performance. The total error detectable by RPT‐QC was determined using EZRULES 3 software.ResultsIn all, 20‐40 data points were needed for RPT‐QC calculations and validated using 20 additional data points. The calculated limits differed among the network of analyzers. The total error that could be controlled was the same or better than that of the manufacturer's commercially available quality control material using the same analyzer for all measurands except hematocrit, which required a higher total error goal than that proposed by ASVCP guidelines to achieve an acceptable probability of error detection. The challenges designed to mimic unstable system performance were successfully detected as out‐of‐control QC.ConclusionsThe challenges for RPT‐QC resulted in acceptable detection of potential unstable system performance. This initial study demonstrates that RPT‐QC limits differ among the network of Sysmex XT‐2000iV analyzers, indicating a requirement to customize for the individual analyzer and laboratory conditions. RPT‐QC could achieve ASVCP total allowable error goals for RBC, HGB, and WBC, but not for HCT. Sigma metrics were consistently >5.5 for RBC, HGB, and WBC, but not for HCT.